Stabilizing Excavation Bottom at Water Table

[theCorkster]

Hi:

I'm currently looking at some alternatives to address loose alluvial soils (silty sand: SM, poorly graded sand: SP) that extend to the water table (-10 bsg). In order to address static settlements above the water table, one approach proposed is to remove and reconstruct to the standards for engineered fill.

The obvious approach would be to dewater the excavation to provide a stable grade in order to build the fill. However, I suspect that given the size of the excavation (about 1 acre), dewatering will be very expensive, and may impact adjacent improvements.

I've also stabilized some soil excavations with cobble (walked into the unstable soil) and geogrid and coarse crushed rock 2 feet).

I'm curious to see if anyone out there might have some experience with this or some other non-structural ( rammed aggregate piers, etc.) approaches to this construction.

 

[SlideRuleEra]

Where I work, coastal SC, the water table is a foot or less below ground. Just excavate, dewatering as necessary with diaphragm pumps, well points, etc., and put in "enough" crushed stone to create a stable working base. Local conditions will determine how much is "enough" - just keep adding more stone as necessary. Once you have this base, then you can begin to add and compact specified materials.

http://www.SlideRuleEra.net

 

[Gimbli]

What type of structure is being proposed? What is the thickness of the loose sands? There are many other alternatives such as Geopiers, Vibroflotation or stone columns, jet grouting, among others. All depends on type of structure and susboil conditions, as always...

 

[theCorkster]

Thanks for the responses.

We've reviewed structural alternative including geopiers, vibro-flotation, stone columns, grouting, and deep foundations w/ structural slabs, etc., and these approaches are viewed as more costly to implement. The over-ex and replacement approach is looked at as the least costly to implement - provided stabilizing the excavation bottom doesn't turn into a bottomless pit.

The sands extend to between 10 to 17 feet bsg, with groundwater at 9 to 10 feet bsg. The client understands that over-ex and replace above water table will not resolve potential seismic settlement of loose sands below groundwater.

 

[DRC1]

Another option is Deep Dynamic Compaction. A wieght is dropped from a given height repeatedly, creating a crater which is backfilled nd compacted by again dropping the weight. Depth of imporvement can be as deep as 20 feet.
I have been on projects to improve bearing on shallow residual soils overlying very punky karst, but I know it has been used for a wide range of poor soil sites.

 

[mudman54]

for a site 1 acre you might be talking dynamic deep compaction dropping weights from cranes.
otherwise, if you can use explosives there have been some loose soils densified with charges.

 

[Mickney]

I am unclear as to the magnitude of the future construction. However, one approach not mentioned is surcharging the site. Loose sandy soils should settle out rather quickly.

 

[dgillette]

I'm not sure surcharging would accomplish much with granular soils, like the SP mentioned, which aren't very compressible even if loose. Back when I worked in FL, we used it for highway embankments over karst features filled with clay and organics, and over clay tailings. It does work for limiting settlements in clay.

If the adjacent improvements are really adjacent, either DDC or explosives is likely to be trouble - even if you don't actually damage the structures, every pre-existing crack or blemish will become yours. I think you should be thinking of stone columns, displacement piles, etc. We've had good luck with jet grouting of silt, but my hunch is the displacement-vibration methods (stone columns, piles, etc.) would be cheaper. (We couldn't use them for the silt because of the low permeability.)

There is also a device that intrigues me for relatively shallow deposits, but I have never seen any hard data on its performance:

http://www.rapidimpact.ca/index.htm

It is small enough that you might need to excavate part way to the water table for the compaction to reach the bottom of the deposit. Don't know a lot of specifics about it, including whether it is easily available outside of Canada (or maybe you are in Canada so it doesn't matter).

DRG

 

[Mickney]

The sand deposit in question is termed a loose alluvial sand; i.e. unconsolidated medium deposited by water. I had a "similar" job here in the piedmont where we partially undercut the alluvial material and surcharged with a soil fill roughly equivalent to the weight of the future construction. Monitored the settlement with plates and when 90% of primarly was believed to have been achieved, removed the surcharge and built the structure. We had the luxury of a larger site so importing/exporting dirt was not an issue. The future construction was a simple two-story wood frame residence as part of a larger residential tract. Not much load at all. So far, no problems to date (5 years and still counting). I gave this option as a consideration; more site specific information would be needed (soil boring data, groundwater fluctuations, foundation data, tolerable settlement limits, etc.) to determine the merits of this option.

 

[dgillette]

The Corkster mentioned seismic concerns with the loose sands. Static loading won't take care of that because it can't change the void ratio of sand nearly enough to change liquefiable to nonliquefiable. That generally requires vibration.

Mickney - What soils were involved at your site? SP would consolidate about as fast as you can place the fill, as would many SMs.

 

[BigHarvey]

Rapid impact Compaction is a ground improvement technique which belongs more or less to the dynamic compaction family.
Whereas in dynamic compaction you have a 15 t weight dropped from 25 m once evry minute, for RIC you use a hydraulic hammer with a ram of 9 tons dropped from 1.2 m 50 times per minutes. The total energy given to the soil will be about the same but since the unit energy is lot less for RIC, the depth of influence will also be much less.
It will be from 3 to 5 meters ( the less fines, the dreater the depth of influence )
I can't imagine that soil replacement could cost less than mass ground improvement unless you completely underestimate difficulties coming along with the presence of groundwater.
Techniques related with dynamic compaction, RIC should prove the more economical.

 

[theCorkster]

Hey, thanks for the great input.

I agree that dynamic compaction would be the most cost effective to improve the soils and reduce the liquefaction potential. One concern is that the site surface will have to be improved - dynamic compaction needs a stiff/dense surface in order to transmit the shear waves to the deeper soils.

I've obtained some information on rapid impact compaction and it in general appears applicable to the site. Since the depth of improvement of this method appears to be up to 20 feet, it seems to fit this site. We do have some silty sands above the water table, and there is some concern as how well this will work (these soils fall between the general zones that respond and do not respond to dynamic compaction).

Fortunately, the sands generally become cleaner with depth, transitioning to poorly graded with silt and poorly graded at and beneath beneath the water table.

Blasting might work as well to densify the soils. Unfortunately, the regulatory aspects associated with this will make other, more conventional approaches seem simple.

 

[Mickney]

Just a quick question on the feasibility of "dynamic" improvement on the site: Are the pore water pressures a concern during dynamic improvements? I typically stay away from dynamic soil improvements in high soil moisture profiles and shallow groundwater conditions. Am I missing something.

 

[dgillette]

Any sort of dynamic improvement (deep dynamic compaction, rapid impact compaction, stone columns, etc.) would generate excess pore-water pressure in the process of densifying the soil, if it is saturated at the time. The structure of the soil is disrupted and the sand is densified, so the water has to go somewhere. There is often a need for a gravel working pad so the water doesn't make a mess of the surface. In granular material, the EPWP dissipates quickly and is localized to the immediate area of treatment (except with blast densification of a large area). You'd have to watch it if you were working immediately at the toe of a steep cut or adjacent to a heavily loaded spread footing, but we've used DDC for seismic modifications of a couple of dams without having any stability problems in the cut slopes.

On the subject of EPWP and blast densification, go to the following site and watch the video for a rather spectacular example in tailings.

http://www.rapidimpact.ca/inco_tailings_pond.htm

 

[BigHarvey]

The original question was to adress settlements of the materials above water table. Dynamic compaction related techniques are well adapted for these materials even if there are some fines in it , under the water table it is another story !
Do not forget that limits of these techniques in terms of fine contents can be improved by combining them with vertical drains like shown in the video link provided by dgillette.